Commentary: AI agents need middle managers too

In the early days of AI, two management scholars predicted that computers would replace middle management by the 1980s. However, they were only partially correct. While AI has automated many tasks previously performed by managers, it has not eliminated the need for human leadership. In the early 2000s, Silicon Valley attempted to flatten organizational structures, but managers quickly returned to their roles due to the complexity of running a large organization.

When the cost of managing corporate information decreases, middle management roles may become obsolete. However, a new structure is emerging with a shorter chain of command, where executives are supported by directly responsible individuals (DRIs) who own the outcomes of their work. DRIs could be middle managers who adapt quickly to AI tools.

As AI takes over routine tasks, middle managers will have more time to exercise judgement, mentor employees, and navigate complex relationships. Human qualities such as trust building, relationship fostering, and guiding teams through uncertainty will become increasingly valuable. Managers will shift from overseeing execution to building trust, fostering relationships, and guiding teams through uncertainty. These skills will be more important than ever as companies rely more on AI agents to execute work.